Everything that changed in Denver local rankings this year, and the four things that did not.

The year in one table

The factors study: what gained and what lost

Whitespark’s Local Search Ranking Factors is the long-running expert survey of the field; the 2026 edition, published November 6, 2025 by Darren Shaw, asked 47 practitioners to score 187 factors from 0 to 5 across the local pack, local organic, conversion and, new this year, AI search visibility. Its chart weightings live in images rather than text, so we will not quote percentages we cannot verify; the directional findings in the report’s own words are enough.

Gained: review signals and behavioral signals. The report says both grew in importance for the local pack. Behavioral means what happens after a searcher sees you: clicks to call, direction requests, website visits, and whether the searcher comes back to the results and picks someone else. Reviews means count, rating, velocity, recency and, increasingly, the words in them, which feed Google’s AI review summaries that BrightLocal says 82% of consumers now read.

Lost ground: on-page and link signals dipped slightly for the pack, and citation signals lost about half a percentage point. That last one gets misread. Citations (consistent name, address and phone across directories) did not stop mattering; they stopped differentiating, because almost everyone competitive has them right. A missing or inconsistent citation still hurts. An extra hundred of them no longer helps the pack.

New: an AI search visibility factor set. The three top-scored factors were presence on expert-curated best-of and similar lists (179), a dedicated page for each service (170), and prominence on key industry-relevant domains (167). Three of the top five are citation-type factors in the broad sense, which is the basis for the report’s line that in AI SEO “mentions (citations) are the new link.” Note the irony: the same word, citation, is slightly down for the map pack and sharply up for the assistants, because the assistants read what the lists and industry sites say about you, not just whether your phone number matches. Our sister directory USA Business Search exists partly for this reason: a complete, verified listing on a directory that AI crawlers can read is a mention with facts attached, not a line in a database.

For the profile-level detail, our Google Business Profile ranking factors for Denver guide is the companion piece.

Two core updates and a spam update

Google confirmed core updates in March and May of 2026, and a spam update that ran from September 24 to October 8. Google’s own statements about core updates are, as always, general: they reassess content quality broadly and are not targeted at local results. The specific local claims below come from vendor analyses, and we flag them as such.

March 2026. Several vendor write-ups reported tighter proximity weighting in competitive local categories after the March update: businesses ranking in suburbs where they had no address lost ground, and the pack contracted toward the searcher. We saw it in Denver for a handful of trades (plumbing and HVAC most visibly), where city-address companies that had been appearing in Aurora and Littleton packs dropped out of them. We did not see it for dentists or law firms, where the pack had already been tight. Treat the mechanism as plausible and the generalization as unproven.

May 2026. One vendor analysis of 8.16 million search results reported organic rankings about 48.7% more volatile from the week of May 18, with mobile results and geo-modified keywords (“roofer denver,” “dentist lakewood”) hit hardest. That is a single study, and volatility is not the same as loss; it means positions moved more than usual in both directions while Google settled. If your “[service] denver” organic ranking swung in late May and settled by July, this is the likely reason, and the right response was to do nothing dramatic.

September 24 to October 8. The spam update is the one that matters most for the way Denver sites are often built. It targets spam policy violations, and the local patterns most at risk are thin location pages (one template, 30 suburbs, the city name swapped) and doorway-style pages built for search rather than people. We published a checklist for local business sites the week it finished, and a separate test for whether your location pages are doorway pages. If your rankings dropped in that window and you have a folder of near-identical suburb pages, start there.

Our own site lost Denver head terms in early October while staying indexed. The cause was not an update; it was 200,000 followed links from our own directories. We fixed that on October 17 and are watching.

A note on reading your own data. Search Console’s Performance report folds AI Overview and AI Mode clicks into Web data, so a drop in clicks with stable impressions can be AI Overviews absorbing clicks rather than a ranking loss. Check average position before you diagnose a penalty. We laid out the setup in how to track AI search traffic and leads.

AI Overviews on local searches: 8% or 68%?

Here is the number everyone quotes and nobody reconciles. Depending on which 2026 study you read, AI Overviews appear on 8%, 21%, 27% or 68% of local searches. All of those can be true at once, because they measure different things.

• Around 8%: independent SERP-tracking studies through the first half of 2026, as summarized by ThinkProfits, for queries with explicit local intent: “plumber near me,” “dentist in Lakewood,” a business name. These are the searches where Google already knows the user wants a map and a phone number, and the pack still leads; one summary put the pack above AI Overviews on these queries by a wide margin.
• 21% to 27%: compilations that count local commercial queries more broadly (“how much does a furnace replacement cost,” “best time to reseed a lawn in Colorado”), where an informational answer fits and Google increasingly writes one. One compilation puts the Q2 2026 figure at 27%, up from 8% a year earlier.
• 68%: Whitespark’s case study on the prevalence of AI Overviews in local search, which tested business-type queries and found AI Overviews on 68% of them, with only 39% returning a local pack. The query set skews toward the kinds of questions where Google answers first and maps second, and the study’s date and method are not fully visible in the published summary, so we treat the number as an upper bound rather than a market average.

What that means for a Denver business is simple once the studies are sorted by query type. For the searches that produce calls today (trade plus city, trade plus “near me,” your name), the map pack still owns the page, and the 2026 study of 500 home-services searches we wrote up in local pack vs LSAs vs Google Ads found the pack on 97.6% of them. For the searches that happen a week earlier, when the customer is still asking what a job costs or whether they need it, AI Overviews now answer first, and the businesses named in those answers are the ones with a real page on the subject. That is the on-page factor Whitespark scored at 170 for AI visibility and slightly down for the pack, and it is why we now treat the two as one engine rather than two budgets; our Denver AI Overviews optimization page describes the work.

The AI local pack test

Through 2026 Google has been testing what the industry calls an AI local pack: a set of provider recommendations, often three, generated inside an AI Overview or an AI Mode answer rather than pulled from the classic map pack. Reports from rater8 and OnPurpose Media describe listings with a rating, a short justification (“praised for same-day service and clear pricing”) and a link, and note that the three businesses in the AI local pack are not always the three in the map pack below it.

Two things about that are important for Denver. The first is that the justification text is built from reviews and your own pages, which means the specific words customers use in reviews, and the specific claims you make on service pages, are now the copy in your listing. A hundred reviews that say “great service” produce an AI local pack entry that says great service; thirty that mention the technician by name, the price being what was quoted and the job being done the same day produce an entry a customer can act on. The second is that the selection draws on the AI visibility factors, not just the pack factors, so a business with strong reviews but no real service pages can hold the map pack and miss the AI pack above it. We have seen exactly that split in Denver for “denver seo company” itself: as of October 19, 2026, the three businesses in the map pack had 80, 64 and 72 reviews at 5.0, and the AI-generated answers above them named a partly different set.

Google has not published how the AI local pack chooses, and the test has appeared and disappeared through the year. Build for it anyway, because the inputs are the same ones that produce calls. The cheapest way to watch it is to search your own trade plus your suburb from a logged-out phone once a week and screenshot whatever sits above the map; if an AI block names three businesses and you are not one of them, that screenshot is your to-do list.

What changed inside Google Business Profile

The profile is where most Denver owners spend their local SEO time, and it changed more in the last two years than in the five before. The removals: chat and call history (gone in 2024), the Q&A API (shut down, so third-party tools can no longer post or manage questions), and a number of the older insights views. The additions: Ask Maps (Gemini answering questions about places from reviews and your profile), Google’s automated calls (which phone you to confirm hours and, since 2025, to ask about availability and pricing on a customer’s behalf), and review-spike pauses, where Google holds new reviews on a profile that gets a sudden burst while it checks them. We checked every change against Google’s own help pages in Google Business Profile changes 2026, with a twelve-point profile check.

Two of those matter for rankings in a way that is easy to miss. Review-spike pauses mean a review drive that lands thirty reviews in a weekend can get them all held, and sometimes removed; the right cadence is one or two a week, forever. And the automated calls now feed Google’s own data about you: if the call asks whether you do water heater replacements on Saturdays and a part-timer says “I don’t think so,” that is now what Google believes. Our guide to answering Google’s AI when it calls has the script.

Google also reported removing or blocking more than 292 million policy-violating reviews in 2025, along with 79 million bad edits and 13 million fake profiles. In Denver trades where fake profiles were common (locksmiths, garage doors, towing), the pack got cleaner this year, which is good news for a real business that simply has not asked for reviews.

The four things that did not move

Distance. Google’s help page still lists relevance, distance and prominence, in that order, and distance is the one no vendor can sell you around. A business at 1001 Bannock Street ranks in central Denver packs and does not rank in Castle Rock packs, and the March update, if the vendor reports are right, made that slightly more true. The honest options are a real second location, a service-area strategy that targets organic and AI answers (which are not distance-limited) in the suburbs you drive to, or accepting the geography. We wrote up the trade-offs in expanding to a second city.

Denver makes distance unusually visible because the metro is a ring of real cities with their own downtowns, not one city with neighborhoods. A searcher in Olde Town Arvada, in Belmar in Lakewood, in downtown Littleton and on Parker Road in Parker each sees a different pack for “plumber,” and the businesses in each one are mostly within a few miles of the searcher. The I-25 corridor compresses this further: Englewood, Centennial and Lone Tree are a fifteen-minute drive apart and three different packs. The practical reading of the 2026 reports is that the rings got slightly tighter, not that they appeared; if you have been planning around a “Denver” pack that covers the whole metro, you have been planning around something that never existed. Our guide to how the Denver suburbs differ for local SEO maps the rings.

The review thresholds. 4.5 stars, 20 reviews then 50, something newer than 90 days, every review answered. BrightLocal moved the consumer numbers in February; Google’s pack and the AI local pack were already rewarding the same things. Nothing in 2026 made a 4.3 with 18 reviews competitive in a Denver trade.

A real page per service. It was the second-highest AI visibility factor in Whitespark’s study, it is what AI Overviews quote, it is what survives a spam update aimed at thin pages, and it is what a booking agent reads for a price. The alternative, one “Services” page listing everything, lost on every front this year.

The count gap. If the three businesses in your pack have 64 to 80 reviews and you have 40, you are not joining them this quarter regardless of what else you do. We say this about our own profile; it has 60 reviews, the pack has 64 to 80, and we are closing it the slow way. Reviews are not the only factor, but the count gap is the one you can most directly close, and it is the first thing we check in a free audit.

A Denver action list by business type

A word on what we changed in our own work this year, since it is the most honest evidence we have. We stopped selling citation volume; consistency is a one-time cleanup, not a monthly line. We moved review work from a quarterly push to a weekly cadence for every client, because of the spike pauses. We rewrote every templated location page we had ever built, ours and clients’, before the September update rather than after it. And we started treating a service page as something an assistant will read aloud, which means a price range, the suburbs by name and a booking link on every one. None of that is clever. All of it is what the year rewarded.

Whatever your row, the measurement is the same: calls and booked jobs by source, not positions. The 2026 changes mostly moved where the first impression happens (an AI answer, an AI local pack, a Google call to your office) rather than what earns it. The businesses that did well in Denver this year were not the ones that reacted to each update; they were the ones with the boring fundamentals in place when the updates arrived, so each one was a tailwind. If you want to know where you stand on them, the free audit checks the pack, the reviews, the pages and the AI answers for your actual market, and you can ask our AI assistant in the header right now what it finds for your business.

Keep going